TWI393453B - Tone detector and method of detecting a tone suitable for a robot - Google Patents

Abstract

A tone detection device for detecting whether an input signal having a tone. The device includes a volume gain calculation unit performing a volume gain treatment on frame data and outputting the volume-gain-treated frame data and energy in time domain thereof, a threshold calculation unit calculating a threshold value on the basis of the energy of the volume-gain-treated frame data, a filter transforming the volume-gain-treated frame data by an algorithm and outputting a characteristic value in a first period, and a comparator comparing the first characteristic value and the threshold value to generate a result and determining the frame data of the input signal has the tone on the basis of the result.

Description

Tone detection method and tone detection device suitable for automatic control device

The present invention relates to a tone detecting apparatus and method suitable for use in an automatic control apparatus, and more particularly to a tone detecting apparatus and method for obtaining an instruction from a tone.

In general, Geze filters are used in telecommunications systems to resolve noise, mute, sound, and tone. For example, U.S. Patent No. 6,381,330 is applied to a telecommunication system, and the receiving end is a line input, and the detection is a Goertzel filter output feature, but the patent is analyzed in a small section in the detection procedure. Instead of the complete time frame analysis, to reduce the amount of computation, and then use the harmonic generation to repeat the determination of the existence of the tone signal. Also, as disclosed in U.S. Patent No. 6,671,252, which is applied to a telecommunication system, the receiving end is a line input, and the detection is performed using an output characteristic of a Göze filter. The detection procedure of this patent uses a short time frame and a long time frame. Different features do different functions to increase the robustness of detection. Short-term frame analysis increases the resolution in the time domain, and long-term frame analysis increases the ability to resist noise. In the conventional telecommunication system, the transmitter and the receiver are connected by a line, and the signal is transmitted to the receiver through the line. If the signal is transmitted wirelessly, the correctness of the detected signal is reduced due to the distance between the transmitter and the receiver. In addition, more specifically, when a speaker emits an audible signal for reception with a microphone that moves relative to the speaker, the correctness of the detection signal is further reduced due to the change in the distance between the speaker and the microphone. .

In view of the above problems, an object of an embodiment of the present invention is to provide a tone detecting apparatus and method capable of increasing detection accuracy. It is an object of an embodiment of the present invention to provide a tone detecting apparatus and method capable of obtaining an instruction from a tone.

An embodiment of the present invention provides a tone detecting device for detecting whether a tone exists in an input signal, and the input signal includes at least one time frame data. The tone detecting device includes: an energy calculating unit and a threshold calculating unit. , a filter and a comparator. The energy calculation unit performs an energy gain process on the time frame data and outputs the time frame data after the energy gain processing and the energy in the time domain of the time frame data, wherein the energy gain process calculates the energy of the frame data in the time domain, and according to The energy of the frame data selectively adjusts the size of the frame data. The threshold value calculation unit calculates a threshold based on the energy of the time frame data after the energy gain processing. The filter converts the energy frame processed time frame data by an algorithm and outputs a first feature value in a first period. The comparator compares the first eigenvalue and the threshold value to generate a comparison result, and judges the presence of the tone in the time frame data of the input signal according to the comparison result.

An embodiment of the present invention provides a tone detection method for detecting whether a tone exists in an input signal, and the input signal includes at least one time frame data. The tone detection method includes: an energy calculation step and a threshold calculation step , a conversion step and a comparison step. The energy calculation step performs an energy gain process on the time frame data and outputs the time frame data after the energy gain process and the energy in the time domain of the time frame data, wherein the energy gain process calculates the energy of the frame data in the time domain, and The size of the frame data is selectively adjusted according to the energy of the time frame data. Threshold calculation step based on energy increase The energy of the time frame data after the treatment is calculated as a threshold. The converting step converts the energy frame processed time frame data by an algorithm and outputs a first feature value in a first period. The comparing step compares the first eigenvalue and the threshold value to generate a comparison result, and judges the presence of the tone in the time frame data of the input signal according to the comparison result.

In an embodiment, the energy calculation unit and the step of the tone detecting device and method further reduce the size of the frame data when the energy of the frame data is greater than a preset value.

In one embodiment, the filter and conversion unit of the tone detecting apparatus and method respectively output a second feature value in a second period, and when the first feature value and the second feature value do not meet a predetermined relationship When the time frame data of the input signal is judged, there is no tone. Preferably, the predetermined relationship is a one-band continuous relationship. And the filter is a Geze filter, the algorithm is a Geze algorithm, and the conversion step is performed by a Geze algorithm.

In one embodiment, the tone detecting apparatus and method further include an instruction output unit and a step, and the instruction output unit and the step respectively determine that the number of groups of the time frame materials in which the tone is present matches a predetermined number of groups, Output an instruction.

The tone detecting apparatus and method according to an embodiment of the present invention can determine the pitch, and in particular, can change the calculation mode according to the change of the relative distance between the receiver and the transmitter to correctly determine the tone.

1 shows a flow chart of a tone detection method in accordance with an embodiment of the present invention. As shown in FIG. 1 , the tone detection method according to an embodiment of the present invention may be adapted to detect whether a tone exists in an input signal Sin. Tone. The input signal Sin contains at least one frame data Fda.

In one embodiment, the input signal Sin can be an audio signal and output by a transmitter that is a loudspeaker and then received by a receiver. The receiver has a function of converting an analog sound signal into a digital sound signal. The input signal Sin may be only a tone Tone, and an instruction is set in the tone Tone according to the frequency and time relationship of the tone Tone, and the frequency of the tone Tone is preferably kept for a different period to set an instruction. In the tone Tone. In this embodiment, a tone Tone having a first frequency and a second frequency is used, and the first frequency is preferably set to be different from the second frequency.

In addition, the input signal Sin, which is a sound signal, may also include a background sound and at least one tone Tone. In actual operation, the background sound can be the background music of the CD player played by the player. Figure 2 is a graph of the minimum loudness (hearing threshold) of the human ear. Please refer to Figure 2. Generally speaking, when the sound signal frequency is 1 kHz, the loudness can be heard as long as 7 db. When the sound signal frequency is 20 kHz, the loudness must be more than 70db for humans to hear. Therefore, the frequency of the background sound can be set to a lower frequency for human perception, and the first frequency and the second frequency of the tone Tone are set to a higher frequency to reduce the situation in which the person is disturbed by the tone Tone while listening to the background music. The frequency of the background sound can be set to less than 16 kHz, preferably less than 10 kHz, and more preferably less than 5 kHz. The first frequency and the second frequency of the tone Tone may be set to a frequency that is less sensitive to the human ear, for example, set to be greater than or equal to 16 kHz, preferably the first frequency is set to 18 kHz and the second frequency is set to 20 kHz.

The following instructions are set to have a start bit and 4 instruction bits. To illustrate, for example, each instruction bit can be a logical value of 0 or a logical value of 1, thus being able to generate 2 quadratic commands. Figure 3 shows a schematic diagram of the time and frequency of an example of a tone. In particular, this tone can contain an instruction with a value of 0101. Table 1 shows an example of an instruction with a value of 0101. Referring to Tables 1 and 3, the first frequency in the tone Tone is continued for a first period t1 (for example, 0.045 seconds) as a start bit. The second frequency is continued for a second period t2 (eg, 0.015 seconds) followed by the first frequency for a third period t3 (eg, 0.03 seconds) as an instruction bit having a logical value of zero. The first frequency is continued for a fourth period t4 (e.g., 0.03 seconds) followed by the second frequency for a fifth period t5 (e.g., 0.015 seconds) as an instruction bit having a logical value of one. The start bit does not have any logical value for identification purposes only.

Figure 4 shows the time and frequency of the tone containing a command bit. intention. An instruction bit b0 contains a plurality of sets of time frame data Fda, and the receiver with an appropriate sampling rate can be selected according to the product design to determine the number of seconds required for each time frame. For example, the instruction bit b0 requires a frequency of 20 kHz for 0.015 seconds and a frequency of 18 kHz for 0.03 seconds for a total of 0.045 seconds. For example, when the number of seconds of each frame of the receiver is 0.045/450 seconds, the instruction bit b0 contains 450 groups. The time frame data Fda, please refer to Table 2 for more detailed description.

Referring to FIG. 1 again, the steps of the tone detection method in this embodiment include: S01: Start.

S10: A signal transmission step of transmitting the input signal Sin to a receiver by a transmitter. In one embodiment the receiver is moveable relative to the transmitter.

S20: an energy calculation step, performing an energy gain processing on the time frame data Fda and outputting the energy in the time domain of the time frame data Fda and the time frame data Fda after the energy gain processing, and calculating the time frame data in the energy gain processing Fda is the energy in the time domain, and selectively adjusts the size of the frame data Fda according to the energy of the time frame data Fda.

S30: The threshold value calculation step calculates a threshold value according to the energy of the time frame data Fda after the energy gain processing.

S40: a conversion step of converting the energy frame processed time frame data Fda by an algorithm and outputting a first feature value in a first period. In an embodiment, a second characteristic value of a second period may be further outputted.

S60: comparing, comparing the first characteristic value and the threshold value to generate a comparison result, and determining, according to the comparison result, that the tone Tone exists in the time frame data Fda of the input signal Sin. In an embodiment, when the first feature value and the second feature value do not meet a predetermined relationship, the time frame data Fda of the input signal Sin is determined to have no tone Tone.

S70: The command judging step outputs an instruction when the number of groups of the time frame materials that are judged to be present in the tone matches a predetermined number of groups.

S90: Waiting for the step, waiting to detect the next set of time frame data Fda. After waiting for a predetermined number of seconds in accordance with the design, the process returns to step S20.

S09: End.

FIG. 5A is a diagram showing an energy meter of a tone detection method according to an embodiment of the present invention. Calculate the flow chart of the first step example. As shown in FIG. 5A, the energy calculation step includes: inputting the time frame data Fda in step S22; and energy gain processing steps. The energy gain processing step includes: step S24 calculates the energy of the frame data Fda in the time domain; step S26 determines whether the energy of the time frame data Fda in the time domain is greater than a value of, for example, a predetermined value Td1 of 29490, and if yes, proceeds to step S28; Otherwise, the energy is outputted to the frame data Fda and the threshold value calculation step S30 is performed. At this time, the time frame data Fda and the energy after the energy gain processing without the step 28 are equal to the time frame data Fda and its energy input in step 22. Step S28: The size of the frame data Fda is reduced by a ratio L1 having a value of, for example, 0.5, and then the flag Cyclecnt of a preset value of 0 is incremented by one (the function will be described later), and the process returns to step S24.

When the Fda energy of the frame data is too large, that is, when the volume is too large, the goertzel filter operation described later may be less accurate, for example, an overflow occurs. Therefore, in this embodiment, the time frame data Fda is first obtained. When the energy is determined and the energy exceeds a predetermined value Td1, the frame data Fda size is reduced by a ratio L1 until the energy of the time frame data Fda after the energy gain processing is less than the predetermined value Td1. More specifically, the predetermined value Td1 can be set to ninety percent of the maximum energy value of a signed word. Furthermore, in an embodiment, an energy calculation unit can be used to perform the energy calculation step.

FIG. 5B is a flow chart showing an example of a threshold value calculation step of the tone detection method according to an embodiment of the present invention. As shown in FIG. 5B, the threshold value calculation step includes: in step S32, it is determined whether the value of the frame data Fda has been adjusted. If the step S34 is performed, the process proceeds to step S36. In detail, when the flag Cyclecnt is greater than 0, the value of the Fda representing the time frame data has been adjusted. For example, refer to FIG. 5A and step S26; if the flag Cyclecnt is equal to the preset value 0, then no. In step S34, the threshold value Td2 is set to the energy of the frame data Fda after the energy gain processing is multiplied by a ratio L2 whose value is, for example, 0.7. In step S36, the threshold value Td2 is set to the energy of the frame data Fda after the energy gain processing is multiplied by a ratio L3 whose value is, for example, 0.48.

When the relative distance between the receiver and the transmitter changes, the volume of the sound signal received by the receiver also changes. The tone detection method of this embodiment dynamically adjusts the threshold value Td2 with the volume received by the receiver. It is possible to more accurately determine whether the input signal Sin has a tone Tone. In addition, in an embodiment, a threshold value calculation unit may be used to perform the threshold value calculation step.

The converting step can be performed in the Geze filter, which uses the Geeze algorithm to convert the time frame data Fda to obtain a first characteristic value in a first period. FIG. 5C is a flow chart showing an example of an algorithm in the conversion step of the tone detection method according to an embodiment of the present invention. The detail shows a flowchart of a Geeze algorithm. Referring to FIG. 5C, the Gezer filter calculates a feedback phase sequence s according to an input sequence x[n] according to the operation program shown in FIG. 5C and using the following mathematical formulas (1), (2), and (3). [n], after calculating the output sequence y[n], the obtained feature values are output.

Wherein, N is the number of samples of the frame data, and f _sample is the sampling frequency (in the embodiment, the sampling frequency of one of the receivers (for example, an AD converter) is less than 10 kHz), and f _tone is the pitch frequency (in this case) In the embodiment, 18 kHz and 20 kHz are used, s _k [-1] and s _k [-2] are feedback storag elements of frequency points and their values are all 0, n is 0, 1, 2 , ..., N-1, s _k [n] is the data of the nth time frame of the kth period, and the eigenvalue y _k [n] is the output of the nth time frame of the kth period of the gerber filter.

The Fourier transform of the discrete signal using the second order recursive infinite impulse response filter by the Geeze algorithm has been commonly used in the field of digital signal processing. It is well known, and a detailed description thereof will be omitted.

FIG. 5D is a flow chart showing an example of a conversion step of the tone detection method according to an embodiment of the present invention. As shown in FIG. 5D, the converting step includes: calculating, in step S42, the eigenvalue y _k [n] of the kth period of the nth time frame and the k+k/5 period of the nth time frame by the Igze filter algorithm, respectively. Characteristic value y _(k+k/5) [n]. Step S44 calculates the difference Diff between y _(k+k/5) [n] and y _k [n] in a predetermined relationship. Step S46 determines whether the difference Diff is greater than 0. If yes, it determines that the frame data Fda is noise, that is, the frame data Fda is regarded as having no tone Tone, and proceeds to step S90; otherwise, the frame data Fda is determined to be working. The signal is sent to step S62, and the working signal may include a background noise or a non-noisy N signal such as tone Tone. In step S44, a predetermined relationship is a band continuous relationship for determining whether the frequency of the current time frame data Fda and the previous time frame data Fda has continuity. In this embodiment, y _(k+k ) is obtained by the following mathematical formula (4). _/5) The difference between [n] and y _k [n] Diff.

Among them, the scale is set to 0.4. When the tone detection method is in operation, there may be noise N in the environment, which affects the correctness of the detection. Since the frequency band of the noise N usually has no continuous relationship and the working signal has a frequency band continuous relationship, it can be measured by When the frequency band Fda of the input signal Sin does not have a continuous relationship, it is regarded as noise and filtered, and the correctness of the detection can be increased. It should be understood that determining whether the input signal Sin has a band continuous relationship is such that the tone detection method can further increase the accuracy of the detection. In an embodiment, it can be determined by the comparison step described later whether the frame data Fda has a tone Tone.

FIG. 5E is a flow chart showing an example of a comparison step of the tone detecting method according to an embodiment of the present invention. The comparison step includes: step S62 comparing the feature value y _k [n] and the threshold value Td2, and determining whether the feature value y _k [n] is greater than the threshold value Td2, and if so, determining that the frame data Fda has the tone Tone and proceeding to step S72; If it is determined otherwise, the frame data Fda has the noise N and proceeds to step S90. In addition, in one embodiment, the tone is determined by comparing the energy in the time domain with the output in the frequency domain, so that the comparison step can be performed using a less expensive comparator to simplify the manufacturing cost.

FIG. 5F is a flow chart showing an example of an instruction output step of the tone detecting method according to an embodiment of the present invention. In an embodiment, an instruction output unit may be used to execute an instruction output step. Preferably, the instruction output unit includes a bit determination unit for performing a bit determination step; and an instruction determination unit for executing the instruction determination step. The bit judging step includes the number of sets of frame data Fda at the time of superimposing each frequency in step S72. In step S74, it is judged whether the number of groups of the frame data Fda at the time of each frequency meets a corresponding predetermined number of groups, and if one bit is output, the process proceeds to step S76; if not, step S72 is performed. Instruction judgment step package The number of various bits is superimposed in step S76. Step S78 determines whether the number of the plurality of bits (for example, the start and/or the command bit) meets a predetermined number of correspondences, if the step S79 is performed; and if not, the step S90 is performed. Step S79 outputs this command.

In detail, please refer to Tables 1 and 2. When it is judged that there is a frequency of 18 kHz, the number of frames of Fda data conforms to 450 groups, and the first bit is output; when it is judged that there is a frequency of 18 kHz, the frame data Fda The number of groups conforms to 300 groups, and it is judged that there is a frequency of 20 kHz when the number of Fda groups of the frame meets 150 groups, and an instruction bit having a logical value of 0 is output; when it is judged that there is a frequency of 18 kHz, the frame data Fda The number of groups conforms to 150 groups, and when it is judged that the number of groups of Fda data having a frequency of 20 kHz conforms to 300 groups, an instruction bit having a logical value of 1 is output. When the number of instruction bits matches 4, the instruction is output.

Obviously, the design of the instruction is not limited by the present invention. For example, an instruction can be set to a frequency of 20 kHz for 0.03 seconds as a logical value of 0, and other frequencies and time are not logical values; or subsequently stop playing. The tone Tone is 0.015 seconds and is regarded as a logical value of 1, so that it can respectively generate 2 first-order commands; or 2 second-order commands. In this case, only the number of time frame data Fda can be calculated. instruction.

Figure 6 is a block diagram showing the function of a tone detecting apparatus according to an embodiment of the present invention. As shown in FIG. 6, the tone detecting apparatus 100 according to an embodiment of the present invention may be adapted to detect whether at least one tone Tone exists in an input signal Sin, and the input signal Sin includes at least one time frame data Fda. The tone detecting device 100 includes an energy calculating unit 110, a threshold value calculating unit 120, a filter 130, and a comparator 140. Also can A transmitter 150 and a receiver 160 are included. Preferably, the instruction output unit 170 is also included.

The energy calculation unit 110 performs an energy gain process on the time frame data Fda and outputs the energy in the time domain of the time frame data Fda and the time frame data Fda after the energy gain processing, and the energy gain processing calculates the frame data Fda in the time domain. Energy, and selectively adjust the size of the frame data Fda according to the energy of the time frame data Fda. The threshold value calculation unit 120 calculates a threshold value based on the energy of the time frame data Fda after the energy gain processing. The filter 130 converts the energy gain processed time frame data Fda by an algorithm and outputs a first characteristic value of a first period. In an embodiment, the filter is a gerber filter, and the algorithm is a zebra algorithm, and the filter 130 can output a second eigenvalue of the second period, and when the first eigenvalue and the first When the two feature values do not conform to a predetermined relationship, the time frame data Fda of the input signal Sin is judged to have no tone Tone. The comparator 140 compares the first characteristic value with the threshold value to generate a comparison result, and judges that the tone Tone exists in the time frame data Fda of the input signal Sin according to the comparison result. The command output unit 170 outputs an instruction when it is determined that the number of sets of the time frame materials in which the tone is present matches a predetermined number of sets. In an embodiment, the instruction output unit 170 includes a bit determination unit 171 and an instruction determination unit 172. The bit judging unit 171 superimposes the number of groups of the frame data Fda at the time of each frequency and judges whether or not the number of sets of the frame data Fda at each frequency conforms to a corresponding predetermined number of sets, and outputs one bit; and the command judging unit 172 superimposes various bits. The number of elements and whether the number of various bits meets a predetermined number of correspondences, and outputs an instruction.

More specifically, the energy calculation unit 110 and the threshold value calculation unit 120, the filter 130 and the comparator 140 can be implemented by a program and a digital signal processor, or can be implemented by a properly designed chip, and the implementation manner thereof will not be described in detail.

The transmitter 150 is a speaker and can output an input signal Sin which is a sound signal. Receiver 160 can be spaced a distance from transmitter 150 and can be moved relative to transmitter 150, which is a loudspeaker, adapted to receive input signal Sin. The receiver 160 can include a microphone 161, an amplifier 162, a narrow wave filter 163, and an analog to digital converter 164. The microphone 161 can be an Electric Condenser Microphone (ECM) for receiving the input signal Sin. Since the intensity of the input signal Sin received by the microphone 161 is sometimes too small or the output signal of the microphone 161 itself is too small, the output signal Sin that the microphone 161 receives and then outputs is amplified. Since the sound and tone emitted by the transmitter 150, which is a speaker, are analog signals, the energy calculation unit 110 implemented by the digital signal processor processes the digital signal, so that it can be set between the microphone 161 and the energy calculation unit 110. The digital analog converter 164 converts the sound signal that is an analog signal into a sound signal of a digital signal. Further, a narrow wave filter 163 may be added between the amplifier 162 and the digital analog converter 164. In an embodiment of the present invention, the frequency of the tone Tone can be set to be 16 kHz or more; and the background sound (the sound including the environment of the tone detecting device 100) is generally 16 kHz or less, so that the narrow wave filter 163 can be used to filter the filter in advance. In addition to the background sound, the device for increasing the tone detecting device 100 and the tone detecting method of an embodiment of the present invention detects the accuracy of the tone.

The method of the embodiment of the present invention, or a specific type or part thereof, may be included in a physical medium such as a floppy disk, a CD, a hard disk, or any other machine (such as a computer). The storage medium is read, wherein when the code is loaded and executed by a machine, such as a computer, the machine becomes a device for participating in the present invention. The method and apparatus of the present invention can also be transmitted in a code format through some transmission medium such as a wire or cable, an optical fiber, or any transmission type, wherein the code is received, loaded, and executed by a machine such as a computer. At this time, the machine becomes a device for participating in the present invention. When implemented in a general purpose processor, the code in conjunction with the processor provides a unique means of operation similar to application specific logic.

The above is only the preferred embodiment of the present invention, and the scope of the invention is not limited thereto, that is, the simple equivalent changes and modifications made by the scope of the invention and the description of the invention are All remain within the scope of the invention patent. In addition, any of the objects or advantages or features of the present invention are not required to be achieved by any embodiment or application of the invention. In addition, the abstract sections and headings are only used to assist in the search of patent documents and are not intended to limit the scope of the invention.

100‧‧‧tone detection device

110‧‧‧ energy calculation unit

120‧‧‧ threshold calculation unit

130‧‧‧ Filter

140‧‧‧ Comparator

150‧‧‧transmitter

160‧‧‧ Receiver

161‧‧‧Microphone

162‧‧Amplifier

163‧‧‧narrow wave filter

164‧‧‧Digital Analog Converter

170‧‧‧Command output unit

171‧‧‧ bit judgment unit

172‧‧‧Instruction judgment unit

1 shows a flow chart of a tone detection method in accordance with an embodiment of the present invention.

Figure 2 is a graph of the minimum loudness of the human ear.

Figure 3 shows a schematic diagram of the time and frequency of an example of a tone.

Figure 4 shows a schematic diagram of the time and frequency of a tone containing an instruction bit.

5A-5F are flow charts showing an example of one of the steps of a tone detecting method in accordance with an embodiment of the present invention.

Figure 6 is a block diagram showing the function of a tone detecting apparatus according to an embodiment of the present invention.

Claims (19)

A tone detecting device is configured to detect whether a tone exists in an input signal, the input signal includes at least one time frame data, and the tone detecting device comprises: an energy calculating unit, performing an energy gain processing on the time frame data And outputting the energy frame processing time and the energy in the time domain of the time frame data, wherein the energy gain processing calculates the energy of the time frame data in the time domain, and selects energy according to the time frame data Adjusting the size of the time frame data; a threshold value calculation unit calculates a threshold according to the energy of the time frame data processed by the energy gain; and a filter converts the energy gain processed by an algorithm The time frame data is outputted to a first characteristic value of a first period; and a comparator compares the first characteristic value and the threshold value to generate a comparison result, and determines the time frame of the input signal according to the comparison result The tone is present in the data, wherein the filter is further outputted to a second feature value in a second period, and when the first feature value and the second feature value do not match When a predetermined relation, determines that the input signal when the tone block information does not exist. The tone detecting device of claim 1, further comprising a transmitter and a receiver moving relative to the transmitter, and the input signal is transmitted by the transmitter to the receiver. A tone detecting device according to claim 1, wherein The energy calculation unit reduces the size of the time frame data when the energy of the frame data is greater than a preset value. The tone detecting device of claim 3, wherein, when the energy of the frame data is greater than the preset value, the threshold value calculating unit multiplies the energy of the time frame data after the energy gain processing Calculating the threshold value by a first ratio; when the energy of the frame data is less than the preset value, the threshold value calculation unit multiplies the energy of the time frame data after the energy gain processing by a different value from the threshold value The second ratio of the first ratio calculates the threshold. The tone detecting device of claim 1, wherein the predetermined relationship is a one-band continuous relationship. The tone detecting device of claim 1, wherein the filter is a gerber filter, and the algorithm is a gaze algorithm. The tone detecting device of claim 1, further comprising an instruction output unit that outputs an instruction when the number of groups of the time frame data that is determined to be present in the tone matches a predetermined number of groups. The tone detecting device of claim 1, wherein the tone comprises a first frequency and a second frequency, the input signal comprising an instruction consisting of a start bit and at least one instruction bit And the tone detecting device further comprises an instruction output unit, wherein the command output unit comprises: a one-bit determining unit, when the number of groups of the time frames of the tone of the first frequency is determined to be consistent a first predetermined number of groups, outputting a start bit; when it is determined that the tone of the first frequency is present The number of groups of frame data conforms to a second predetermined number of groups, and when it is determined that the number of groups of the time frame materials of the tone having the second frequency meets a third predetermined number of groups, the output one has a first logical value. An instruction bit; when the number of sets of the time frame data of the tone at which the first frequency is determined to exist is in accordance with a fourth predetermined number of sets, and the time when the tone of the second frequency is determined to be present When the number of groups of the frame data meets a fifth predetermined number of groups, outputting an instruction bit having a second logical value; and an instruction determining unit, when the number of the at least one instruction bit meets a predetermined number of bits , output the instruction. The tone detecting device of claim 8, wherein the second predetermined number of groups is equal to the fifth predetermined number of groups, the third predetermined number of groups is equal to the fourth predetermined number of groups, and the first predetermined number is The number of groups is equal to the second predetermined number of groups plus the third predetermined number of groups. The tone detecting device of claim 8, wherein the first frequency and the second frequency are both greater than or equal to about 16 kHz. A tone detection method for an automatic control device for detecting whether a tone exists in an input signal, the input signal includes at least one time frame data, and the tone detection method includes: an energy calculation step, the time frame The data is subjected to an energy gain process and outputs the time frame data of the energy gain process and the energy in the time domain of the time frame data, wherein the energy gain process calculates the energy of the time frame data in the time domain, and according to The energy of the frame data selectively adjusts the size of the time frame data; a threshold value calculation step calculates a threshold value according to the energy of the time frame data after the energy gain processing; a conversion step of converting the energy gain processed time frame data by an algorithm and outputting a first feature value in a first period; and a comparing step of comparing the first feature value with the threshold value to generate a Comparing the result, and determining, according to the comparison result, that the tone is present in the time frame data of the input signal, wherein the converting step is further outputting a second feature value in a second period, and determining the first feature value and the When the second feature value does not conform to a predetermined relationship, it is determined that the tone of the time frame data of the input signal does not exist. The tone detection method of claim 11, further comprising transmitting, by a transmitter, the input signal to a receiver that moves relative to the transmitter. The tone detecting method of claim 11, wherein the energy calculating step reduces the size of the time frame data when the energy of the time frame data is greater than a preset value. The tone detecting method of claim 13, wherein the threshold value calculating step determines, when the energy of the time frame data is greater than the preset value, multiplying the energy of the time frame data after the energy gain processing Calculating the threshold value by a first ratio; determining that the energy of the time frame data after the energy gain processing is greater than the preset value, multiplying the energy of the time frame data after the energy gain processing by a difference from the first ratio The second ratio calculates the threshold. The tone detecting method of claim 11, wherein the predetermined relationship is a one-band continuous relationship. The tone detecting method according to claim 11, wherein the converting step is performed by using a Geze algorithm. For example, the tone detection method described in claim 11 is applicable. There is further included an instruction output step of outputting an instruction when the number of sets of the time frame data of the tone is determined to be a predetermined number of groups. The method for detecting a tone according to claim 11 further includes an instruction determining step, the step of determining the command comprising: a one-bit determining step, when it is determined that there is a time frame of the tone of the first frequency When the number of sets of data conforms to a first predetermined number of groups, a start bit is output; when it is determined that the number of sets of the time frame data of the tone of the first frequency is consistent with a second predetermined number of groups, Determining that the number of groups of the time frame materials having the tone of the second frequency conforms to a third predetermined number of groups, outputting an instruction bit having the first logic value; when it is determined that the first frequency exists The number of sets of the time frame data of the tone conforms to a fourth predetermined number of groups, and when it is determined that the number of groups of the time frame materials of the tone having the second frequency meets a fifth predetermined number of groups, output one An instruction bit having a second logic value; and an instruction determining step of outputting the instruction when the number of the at least one instruction bit matches a predetermined number of bits. The tone detecting method of claim 18, wherein the second predetermined number of groups is equal to the fifth predetermined number of groups, the third predetermined number of groups is equal to the fourth predetermined number of groups, and the first predetermined number is The number of groups is equal to the second predetermined number of groups plus the third predetermined number of groups.